A great deal of research has been devoted to understanding the functions of forebrain acetylcholine (ACh) in mammalian cognitive processes. However, conclusions from many of the studies have been tempered by caveats related to the inherent limitations of pharmacological and lesion approaches. Thus, specific functions of ACh remain equivocal. In order to look directly at the role of ACh in learning, we will develop a conditional genetic approach which will allow us to eliminate ACh from discrete regions of the brain with spatial and temporal precision. Specifically, mice carrying a loxP modification of the gene encoding choline acetyltransferase (ChAT) will be generated by standard gene targeting techniques. Adult homozygous mice will be administered a viral vector encoding Cre recombinase. Infected neurons will functionally inactivate their ChAT genes, and thus will no longer be able to synthesize ACh. Once the parameters of the methodology are worked out, we will determine if loss of basal forebrain ACh alters either basic behavioral or cognitive functions. If successful, this system will have general applicability for re-examination of many putative functions of ACh in mammals.